CoMind Early Feasibility Study

Recruiting

• Conditions: Intracranial Pressure Changes; Intracranial Pressure; Intracerebral Hemorrhage; Encephalopathy; Hydrocephalus; Encephalitis; Stroke; Autoregulation; Traumatic Brain Injury

• Registration Number: NCT06368648
• Lead Sponsor: CoMind Technologies Limited

Brief Summary

The purpose of this research, which has been determined as non-significant risk by the central IRB overseeing the study, is to obtain information to help further develop a machine (a medical device) to measure the pressure around the brain from the outside (this pressure is called intracranial pressure or ICP). Monitoring and managing ICP is an important part of care for patients with conditions such as Traumatic Brain Injury (TBI). However, the current way of measuring ICP requires surgery to drill a hole into the skull, and therefore can introduce additional risks such as infections and pain.

Recent research has shown it may be possible to measure ICP without needing surgery. This technology is in development, but large amounts of data is required to build these new devices.

Through collecting a large database of information from patients who have both the routine surgical device and the research device applied to their head, the research team will work to develop and test an effective and potentially safer way of monitoring patient ICP.

Detailed Description

A prospective, observational study, determined as non-significant risk by the Central IRB, to assess and improve performance of a non-invasive ICP estimation system. The development of non-invasive systems is intended to replace the need for invasive ICP monitors. Participants serve as their own controls with concurrent, synchronous measurements of ICP (measured invasively as per standard clinical practice), Arterial Blood Pressure (ABP- measured invasively as per standard clinical practice), and measurements of a novel non- invasive cerebral blood-flow index (CBFi) from the CoMind One EFS device. These signals will be recorded simultaneously. The ABP and CBFi will serve as inputs to a model which outputs a non-invasive estimate of ICP, and the invasive ICP signal will be used to supervise and evaluate the performance of the non-invasive ICP estimation model. A large volume of data shall be collected for the purposes of training and testing the non-invasive ICP model. The Sponsor will report limits of agreement (LOA) between data-driven estimates of ICP and the invasively measured signal. In addition, this study will compare metrics built on non-invasive estimation of ICP or CBFi that indicate the state of CAR, and the limits of Autoregulation.

Study Timeline

• Study First Submitted: 2024-04-11
• Study First Submitted QC: 2024-04-11
• Study First Posted: 2024-04-16
• Study Start: 2024-11-27
• Status Verified: 2025-01
• Last Update Submitted: 2025-01-14
• Last Update Posted: 2025-01-15
• Primary Completion: 2025-10
• Study Completion: 2025-10

Recruitment & Eligibility

• Status: RECRUITING
• Sex: All
• Target Recruitment: 581

Inclusion Criteria

1. Male or female sex at birth, and aged 18 years or older on the date of enrollment.
2. Receiving continuous invasive ICP monitoring (Bolt or EVD) as part of standard care.
3. Invasive ICP monitor catheter penetrating the parenchyma or ventricles.
4. Receiving continuous invasive ABP monitoring as part of standard care.

Exclusion Criteria

1. Presence of any implant (cosmetic or otherwise) in the frontal bone in such proximity to the CoMind One EFS Sensor that they might physically touch.
2. Open wounds on the forehead such that CoMind One EFS Sensor cannot be safely placed over an area of intact skin
3. Presenting with radiographic evidence of a non-intact skull at the recording site on admission.
4. If patient is enrolled in an intervention/study that may interfere with SoC ICP measurements or the CoMind One EFS device measurement then the patient is ineligible.
5. Patients with decompressive craniectomy will be excluded unless a CoMind One EFS recording can be made from intact skull.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Models trained using CoMind CBFi and ABP will be evaluated based on their limits of agreement (LOA) with invasive ICP.	From enrollment to the end of observation (either 3 weeks after removal of the CoMind device or until ICU discharge, whichever event is sooner)	The threshold limit of agreement will be evaluated at three levels corresponding to a model with clinical utility that is optimal, high, or moderate clinical utility. These levels of clinical utility are defined as follows: ● Optimal clinical utility: a model trained using CoMind CBFi and ABP can achieve LOA with invasive ICP less-than or equal to +/-2 mmHg in the range of ICP between 10-30 mmHg, and less-than or equal to +/-4 mmHg in the ranges of 0-10 mmHg, and 30-50 mmHg. ● High clinical utility: a model trained using CoMind CBFi and ABP can achieve LOA with invasive ICP less-than or equal to +/-3 mmHg in the range of ICP between 10-30 mmHg, and less-than or equal to +/-5 mmHg in the ranges of 0-10 mmHg, and 30-50 mmHg. ● Moderate clinical utility: a model that achieves LOA with invasive ICP less-than or equal to +/-4 mmHg in the range of ICP between 10-30 mmHg, and less-than or equal to +/-6 mmHg in the ranges of 0-10 mmHg, and 30-50 mmHg.
A test of the difference in estimation errors between two non-invasive ICP estimation models: one trained with ABP and CoMind One EFS CBFi, and one trained only using ABP	From enrollment to the end of observation (either 3 weeks after removal of the CoMind device or until ICU discharge, whichever event is sooner)

Secondary Outcome Measures

Name	Time	Method
Demonstration of a non-invasive ICP model that can classify the transition between ICP 'states', i.e. transitions between low ICP (less than 20 mmHg), and high ICP (greater than 20 mmHg).	From enrollment to the end of observation (either 3 weeks after removal of the CoMind device or until ICU discharge, whichever event is sooner)
Demonstration of an non-invasive ICP model that can predict trends in ICP	From enrollment to the end of observation (either 3 weeks after removal of the CoMind device or until ICU discharge, whichever event is sooner)
Demonstration that non-invasive ICP estimation using the CoMind One EFS device is insensitive to differences in skin tone across the population.	From enrollment to the end of observation (either 3 weeks after removal of the CoMind device or until ICU discharge, whichever event is sooner)
Demonstration of a non-invasive ICP waveform prediction model and assessment of its accuracy against the invasively-measured ICP waveform.	From enrollment to the end of observation (either 3 weeks after removal of the CoMind device or until ICU discharge, whichever event is sooner)

Trial Locations

Locations (1)

University of Washington, Harborview; 🇺🇸 Seattle, Washington, United States